search

kezhang-cs / Video-Summarization-with-LSTM

Implementation of our ECCV 2016 Paper (Video Summarization with Long Short-term Memory)
Other
197 stars 54 forks source link

Interpreting the dataset #17

Closed ayushkothari27 closed 3 years ago

ayushkothari27 commented 4 years ago

@kezhang-cs Hello, I see there are two fields in the dataset - gt_1_i and gt_2_i. Can you explain the difference and significance of these? I see that you have used gt_1_i in your code while calculating losses. In the readme file, you have written gt_1_i is the frame-level importance scores while gt_2_i is the binary value transferred from the original dataset. I couldn't understand this part. Thank you.

ritvik2708 commented 4 years ago

Hey @aayushkothari11 I think gt_1_i is groundTruth_1 for ith sample and similarly for gt_2_i. gt_1 gives the outputs for the summary in soft values i.e. grey values b/w 0 and 1 while gt_2 provides a binary values which indicates either the frame is part of gt else it is not.

Chen414 commented 3 years ago

Hello, l also need the datasets, but the link is not available, can you send the datasets to my emails 644109863@qq.com? I'll be very grateful to you~Thank~

Chen414 commented 3 years ago

Hello@ritvik2708, l also need the datasets, but the link is not available, can you send the datasets to my emails 644109863@qq.com? I'll be very grateful to youThank